Somewhat related is how libraries like three.js don't follow semver and only use a single incrementing version number. Reading the changelogs, it also looks lik

Somewhat related is how libraries like three.js don't follow semver and only use a single incrementing version number. Reading the changelogs, it also looks like they never make any mistakes. It seems that there has never been a point in three.js's history where there's been a critical security flaw or bug severe enough to justify incrementing the version number solely for a bugfix. I have no idea how they do it. Personally I would prefer to keep things as simple as possible, but unlike the maintainers of those projects, when it comes to making releases, I'm not perfect.